healthcare delivery

Skeptics – We have Proof that Operational Excellence Works in Healthcare!

I’ve written a number of times on the skepticism and lack of understanding surrounding operational excellence (Op-ex) deployments in healthcare. Under the op-ex umbrella, there are several change management initiatives such as lean, Six Sigma, Total Quality Management, etc. All of these emphasize the importance of a systemic approach toward the sustainable improvement of key performance metrics.

Although there are several books and research on the benefits from these initiatives in healthcare context, skeptics dismiss them as mostly anecdotal and devoid of the rigor necessary for causal inferences (i.e., the presence of op-ex leads to healthcare improvements). In this regard, the skeptics are right. Many existing inferences in this area come from surveys (i.e., asking healthcare leaders about op-ex and correlating results to performance) or are based on secondary data that may not adequately capture op-ex elements.

Recent research Peter Ward & I conducted with Dr. Susan Moffatt-Bruce and other collaborators at The Ohio State University Wexner Medical Center leaves little room for skepticism about the impact of op-ex deployment in healthcare. Our quasi-experimental investigation permitted us to compare care delivery performance before and after deployment, using a control group from the same setting to account for any variations in the system beyond our experiment. Results show that op-ex deployment reduced 30-day readmissions, decreased monthly operating costs, improved patient satisfaction and also improved the quality of work among caregivers. In short, it works.

The Quasi-Experiment

Our experimental intervention was conducted over a period of three years and used op-ex principles to change the way patients were discharged after a kidney transplant. We specifically looked at a fundamental op-ex element: Standard work during patient discharge. This represents the set of activities standardized in terms of content, sequence, timing and outcome but still able to “flex” based on patient characteristics. An important feature of standard work is that it should be designed and developed by people doing the work and continuously monitored for improvement. In our study, more than 40 nurses, physicians, social workers, IT specialists and nurse managers responsible for discharging the patients designed and developed the new standard work over a nine-month period.

We focused on the kidney transplant discharge process because these recipients require prolonged periods of care after discharge. In fact, one in three U.S. kidney transplant patients  is readmitted within 30 days, resulting in higher hospitalization costs, resource utilization (i.e. longer wait times for other wait-listed recipients), and – more importantly – poor quality of life for the patient post-procedure. When looking closely at the discharge process, we found patients were taking in an overwhelming amount of information in the four days prior to discharge. In this short period, caregivers covered more than 45 topics, including medication, infection prevention, lab compliance, fluid intake, exercise, and dental hygiene, among many others. Some of these instructions – dental hygiene, for one – were not critical to patients’ immediate well-being and could have been given post-discharge. Nurses, we found, also varied their delivery of instructions, adding to confusion and anxiety. Figure 1 represents the current state of discharge process prior to our intervention. As seen from this figure, variability during the discharge process confused the patients and heightened anxiety levels after discharge, triggering readmissions.

Figure 1

 

The new standardized discharge work sought to resolve these inconsistencies. The caregiving team developed specific instructions on the content, sequence, timing, and delivery outcome for each element of discharge, drawing from data on 15 former transplant patients. The team also decided to adopt a two-part instructional approach, with inpatient nurses giving the most essential discharge instructions during the hospital stay (Part I Instructions), while outpatient nurses gave other instructions (Part II Instructions) 48 hours after discharge. This resulted in a formal handoff process between the inpatient and outpatient caregiving teams. After implementing this new process, the team also adopted 10-minute “huddles,” or standing meetings, to continuously monitor and improve communication at inpatient (bi-weekly) and outpatient (weekly) units by focusing on problem solving and best practice-sharing. These huddles have been going on for more than a year, with full participation from the caregiving teams. The entire op-ex implementation, launched in May 2015, took about a year. Figure 2 represents the process after the op-ex implementation.

Figure 2

To validate the benefits of the Op-Ex implementation, we tracked the discharge process outcomes for more than 700 kidney transplant patients for four years (two years prior to, and one year after, implementation). To ensure our data weren’t skewed by factors such as discharge procedure changes or electronic medical record implementation, we collected patient discharge process data from 180 heart and liver transplants, using them as a control group. Patients undergoing heart and liver transplants experience similar outpatient issues as kidney transplant recipients and have very similar discharge instructions (as seen in Figure 1). We also controlled for patient factors such as age, preexisting conditions, length of stay, functioning of transplanted organ at the time of discharge, ethnicity, 30-day mortality rates, etc. In addition, we adjusted for any other changes to the process that were not a part of our experiment (e.g. adding new technologies)

After accounting for all these effects, we used a difference-in-difference (DID) approach to estimate the causal effects of our implementation on readmission outcomes. Our analyses suggest that the likelihood of getting readmitted to the hospital was about 35% lower for the treatment group (kidney transplant process after Op-Ex implementation) compared to the control group and pre-implementation group (See Figure 3).  On average, this accounted to a 25% reduction in the overall readmission rates after the intervention for the treatment group, resulting in a conservative savings estimate of $85,000 due to hospitalization expenses for the unit. This doesn’t capture savings from the overall patient well- being after the surgery.

We also found a 10% increase in the overall patient experience quality score, measured using the Hospital Consumer Assessment of Healthcare Providers and Systems (HCAHPS) survey. Interestingly, the control group unit experienced an 11% decline in these scores during the same time period (see Figure 4). Improving the scores from this survey has been a key objective for hospital administrators as Medicare/Medicaid reimbursements are now tied to them.

 

Figure 3

Figure 4

Our study shows that Op-Ex transformation indeed can result in improvement across multiple outcomes. The real challenge is how to initiate and sustain this journey. All these results were possible due to the hard work of all the front-line nurses in this unit. They truly are the miracle workers who continue to sustain this journey and innovate as they move along.

 

Glad that I am not alone talking about HCAHPS!

istock-HCAHPS-survey

 

The Hospital Consumer Assessment of Healthcare Providers and Systems (HCAHPS) survey measures the perception of care delivery from a patient point of view. Centers for Medicare and Medicaid (CMS) have historically used this survey as a measure of healthcare delivery effectiveness. Beginning FY 2013, they started measuring hospital’s performance on HCAHPS and made it accountable as a part of value based purchasing program. I have done research with several colleagues from operations and medicine assessing the relationships between HCAHPS survey outcomes and  hospital’s clinical performance such as 30-day readmissions (See https://hbr.org/2015/09/what-has-the-biggest-impact-on-hospital-readmission-rates).

However, there has always been a skepticism among physicians on the use of this survey instrument. I have listened to so many physicians criticizing this tool and blaming it for subjectivity and other measurement issues (e.g. anonymity, scaling issues etc). While their arguments on the survey instrument has some merits, this is the only standardized tool deployed across US hospitals. My argument has always been that despite these issues, HCAHPS allows hospitals to evaluate their care delivery process (not the people!) and can be used to improve them on a continuous basis. I clearly understand my limitations as a business school professor  (a pseudo Doctor) discussing the benefits of patient-centric healthcare delivery among the real doctors with MDs. Now I am not alone in making this argument.  A recent study in the JAMA Internal Medicine (http://archinte.jamanetwork.com/article.aspx?articleid=2513630) supports this viewpoint and finds that hospitals with better ratings on HCAHPS are associated with reduced 30-day readmission rates and risk adjusted mortality rates. The study co-authored by Ashish Jha and his colleagues analyzed data from 3076 hospitals and finds that hospitals with higher star ratings are associated  (note -not causality!) with lower readmissions and mortality.

The real question is that whether this would trigger some changes among physicians on the utility value of HCAHPS now that their fellow physicians arrive at a similar conclusions that are published top medical journal – only time will tell!

 

 

In Hospitals, a trade-off between patient safety and patient experience

Balance

Hospitals that adopt strategies to reduce errors and meet government requirements face an initial tradeoff between improved patient safety and a decline in the quality of individual patients’ experiences, according to new research.  Quality process management, a practice associated with the private sector, is becoming more common in hospitals as they set up operating systems in response to federal and state mandates to reduce medical errors and improve patient safety, the researchers say. It is not surprising to find that the implementation of these techniques led to improved safety outcomes. But finding that these improvements sometimes came at the expense of the quality of the patient experience was unexpected. Also referred to as experiential quality, the quality of the patient experience is gauged by how patients perceive their personal interactions with health-care providers.  “Patient Safety is about doing things correctly – strict guidelines, standardization and checklists, for example – so when you consider experiential quality is about customizing health-care delivery to an individual patient’s needs, there is a tension there,” said Aravind Chandrasekaran, associate professor of management sciences at Ohio State and lead author of the study.

“When leaders were more patient-centric, our analysis showed that they were able to overcome that tension between patient safety and the quality of the patient experience,” Chandrasekaran said. “Leaders have to be thinking about patients when they design their operations. That way they can cater many of their design principles to individual patients.”

How might this tension play out? Consider a patient  safety guideline of giving a beta blocker prescription for patients who have had a heart attack, but offer no suggestions for how to effectively relay that information to a patient. So the hospital gets a good mark for prescribing the drug, but a patient may not understand the instructions and possibly won’t even fill the prescription.

Chandrasekaran and colleagues assert that setting up standardized quality management systems is the most effective way for hospitals to meet state and federal mandates geared toward patient safety. Quality process management entails a systematic approach to map, improve and adhere to given sets of guidelines with a goal to minimize an organization’s variation in its processes.  Federal and state regulations in health care have become more stringent since 1999, when the Institute of Medicine released a milestone report stating that almost 100,000 people died every year as a result of preventable medical errors in U.S. hospitals.

In a move toward standardization, the U.S. Centers for Medicare and Medicaid Services (CMS) in 2003 issued hospital care guidelines related to four health conditions: heart attack, heart failure, pneumonia and surgical care. CMS requires hospitals to report their care practices with these types of cases, and has provided financial incentives to hospitals that are best at adhering to the standards of care outlined in these guidelines. In addition, many states have passed patient-safety legislation calling for reductions in hospital-acquired infections, also beginning in 2003. In the study, the researchers used this legislation as an example of state leadership focused on improving patient safety.

These regulations have led many hospitals to adopt quality process management practices to improve their safety outcomes as quickly as possible. But the researchers also wanted to examine what happened to the patient experience as hospitals focused on new techniques to improve their clinical quality. To determine these relationships, they analyzed four sources of data: a survey of 284 acute care hospitals in 44 states; CMS patient safety scores publicly reported between April 2009 and March 2010; state legislative mandates for reduced hospital-acquired infections passed between 2003 and 2008 in a portion of those 44 surveyed states; and April 2009-March 2010 reports from the Hospital Consumer Assessment of Healthcare Providers and Systems survey as a measure of patient experience quality. Directors of quality or chief nursing officers at 284 hospitals in 44 states were surveyed to determine how extensively respondents were using a data-driven, quality management system to design operations and train staff with the goal of adhering to CMS guidelines. Additional questions examined the leadership style and culture of each hospital.

The analysis showed that a focus on quality process management was simultaneously associated with an increase in patient safety as reported by hospitals and a decrease in the quality of the patient experience as reported by patients. State legislative mandates to improve patient safety initially reinforced this tradeoff. However, the earlier these laws were passed, the sooner hospital environments adjusted to operational changes so they could improve the patient experience as well.  When analyzing the  survey data about hospital leaders’ traits, we found that patient-focused leadership could soften the negative association between quality process management and experiential quality, allowing hospitals to excel in both areas.

Hospitals also commonly survey patients after their stays to gauge their satisfaction with their care. These surveys collect information about patients’ overall impression of their care and whether they would return to the hospital or recommend its services to friends and family. This suggests that patients are able to understand and appreciate the standardization in hospitals if they experience better levels of communication during their care. In other words, treating patients well enables them to better perceive when they are getting the ‘correct’ clinical care.

Five Star Treatment – Moving the Needle on Patient Safety and Experience

Moving the needle

Whether they like it or not, hospitals today can’t solely measure success by properly treating patients according to established standards of care. Moving a patient through surgery to recovery with no complications or hospital-borne bugs, for example, is only part of the equation in the eyes of the government agencies holding the purse strings for critical Medicare and Medicaid reimbursement dollars. Now, patience experience also matters – and hospitals must wrangle with age-old cultural tensions to avoid getting hit where it hurts.

To avoid penalties and bring in maximum federal reimbursement funds, hospitals since 2013 have had to show high marks in evidence-based standards of clinical care and the Hospital Consumer of Healthcare Providers and Systems, or HCAHPS, survey. This survey measures patients’ experience, driven largely by how they grade their interaction with physicians and nurses. Pressure is mounting, as reimbursement penalties are set to potentially double next year if hospitals don’t show improvement on clinical care and patient experience under terms of Obamacare, the Patient Protection and Affordable Care Act passed in 2010. In addition to these penalties, we also find that improving both these dimensions can actually reduce the occurrence of 30-day readmission rates1.

Physicians greet this newfound emphasis on patient experience with skepticism at best and outright resentment at worse, largely due to the subjective nature of the HCAHPS survey. This, however, perfectly illustrates the challenge therein: It’s difficult to simultaneously improve these two measures in a notoriously hierarchical culture where doctors focus relentlessly on disease and nurses fear challenging them, even while carrying the torch as the patient advocate. Luckily, it’s possible, and our forthcoming research2 points to key measures that can drive the culture change needed to make it happen.

Magnet

The Magnet matters

These findings come five distinct case studies that involve more than 50 interviews we conducted at five large acute-care hospitals. This was paired with data drawn from more than 3,000 hospitals from 2006-12, before and after Obamacare kicked in. Strikingly obvious is the sheer difficulty of improving clinical care and patient experience without trade-offs (also evidenced in a different study3 conducted in 2012). Clinical care is comparably easier to operationalize and roll out through top-down directives: keep quality consistent, keep patients safe, and keep variation during care delivery minimal. The problem? These directives can make little room for the customized patient care in nurses’ hands. At the same time, providing a patient with a Ritz Carlton-quality experience means nothing if it interferes with doctors working to keep them alive and well.

A winning strategy a number of hospitals are beginning to adopt, we found, starts with clinching “Magnet” hospital status through the American Nurses Credentialing Center (ANCC). Its Magnet Recognition Program® is the stamp of a high-quality program, denoting they are involved in decision making from the highest strategic levels to the patient’s bedside. The presence of this “bottom-up” decision making, we found, is a crucial factor in hospitals’ ability to improve on “hard,” evidence-based patient care and the “soft” patient experience side. In fact, we found hospitals with ANCC Magnet status were 24 percent more likely than their peers to show improvement in both of these measures, according to federal data from the Centers for Medicare and Medicaid (CMS).

Flattening this hierarchy doesn’t mean the end of administrative oversight, but it does require a newly measured approach to it. What emerged in our research was a “sweet spot” of administrative duties for nurses. Hospitals whose nurses average more than 25 percent of their shifts completing checklists and managing electronical medical records instead of patient are more likely to show weaker trends in improving on both dimensions. But so are hospitals whose nurses spend scarcely more than 10 percent of their time on the same tasks. Administrative oversight and an empowering culture for nurses can peacefully exist, but only if both are carefully balanced.

Playing matchmaker

To maintain the culture needed to treat patients and treat them well, providers can learn much from the physician-nurse interaction in the five hospitals we closely studied. Collaboration between doctors and nurses at the same rung on the ladder – Chief Nursing Officers and Chief Medical Officers or Medical Director – Nursing Director – is a longstanding best practice, but it’s not enough. In another forthcoming research4 on collaboration, we found that the best hospitals took this a step further, routinely pairing experienced, high-level nursing staff with low-level doctors, aiming to instill the importance of patient experience early in their careers. The reverse happens, too: One hospital, for example, paired a chief medical officer with an entry-level nurse and noted success in breaking down the hierarchy that can prevent nurses from speaking up. Taken together, these cross-level collaborations can drive major gains in the improving both the clinical and patient experience measures that matter most.

These crucial collaborations can take many forms. Hospitals we studied reported placing nursing leaders on physician-hiring committees or sending medical directors to work with nurses on day-to-day tasks. One even borrowed from iconic General Electric CEO Jack Welch, issuing a “get on board or get out” policy for its new collaborative culture. These surely move the needle, but they’re also formal measures that “force” physician-nurse interactions. The most successful hospitals see these collaborations form and grow organically, whether it’s nurses inviting physicians to their meetings or informal touch-base discussions in hallways. These connections might not be in the service of scratching another item off a checklist, but their impact can be just as powerful.

 

Articles

  1. C. Senot, A. Chandrasekaran, P. Ward, A. Tucker, and S. Moffat-Bruce, “The Impact of Combining Conformance Quality and Experiential Quality on Readmissions and Cost Performance,” Management Science (2016).
  2. Senot, C., Chandrasekaran, A., Ward, P. 2016. “Role of Bottom-Up Decision Process in Improving the Quality of Health Care Delivery: A Contingency Perspective“. Forthcoming at Production and Operations Management
  3. A. Chandrasekaran, C. Senot, K. Boyer, “Process Management Impact on Clinical and Experiential Quality: Managing Tensions Between Safe and Patient-Centered Healthcare,” Manufacturing and Service Operations Management 14, no. 4 (2012): 548-566.
  4. C. Senot, A. Chandrasekaran, P. Ward. 2016. Collaboration between Service Professionals during the Delivery of Health Care: Evidence from a Multiple- Case Study in U.S. Hospitals. Forthcoming in Journal of Operations Management. 

 

Do we understand Lean Deployment in Healthcare?

Patient Safety                                             Lean

Anyone who has taught lean principles grounded in the famous Toyota Production System (TPS) to organizations outside the manufacturing industry has – at least once – heard this common refrain: “(Insert industry here) isn’t cars on an assembly line. This doesn’t apply to my work.”

Leading lean thinkers, of course, have learned how to work with individuals and teams to move past this roadblock and garner buy-in – that’s why the practices and tools intrinsic to TPS have made their way into countless industries. Lean still can be a target for criticism, though, and one need look no further for proof than an article published earlier this year in the New England Journal of Medicine – and the debate it ignited.

The January issue of NEJM featured an article called “Medical Taylorism” where authors and physicians Pamela Hartzband and Jerome Groopman assert that lean principles “cannot be applied to many vital aspects of medicine. If patients were cars, we would all be used cars of different years and models …” This tipped off a flurry of rebuttals, including one from Lean Enterprise Institute CEO John Shook boldly titled “Malpractice in the New England Journal of Medicine.” In his piece, which itself attracted widespread attention, Shook writes that the foundational lean principles of continuous improvement and respect for people are critically important in the health-care system.

Shook is right, but I’d like to approach this discussion from a different angle, namely that this line of criticism has emerged elsewhere – and it’s rooted in a lack of understanding of lean deployment.

One of my initial research areas sought to understand how standardization and “smart application” of Design for Six Sigma (DFSS) principles can aid R&D and innovation efforts. I pursued this as a number of business press publications and industry practitioner blogs lamented the damage Six Sigma does to creativity and praised the need for variation for innovation. Several years of research1,2 at Fortune 500 companies with my colleagues in Fortune 500 companies made us realize such sentiments don’t hold much water. We found ambidex, in fact, that principles of Six Sigma/DFSS – when applied to the innovation process correctly (hence the “smart” in “smart application’) – can help reduce unnecessary variation and stop worthless innovation activities that consume R&D funding (shift).

I’ve more recently collaborated with researchers and physicians to tackle similar questions in health care. Once again, the findings – published in several academic and practitioner outlets – are very similar: The smart application of lean and continuous improvement principles can help develop a safe and patient-centric healthcare system.

In arguing that patients aren’t cars, the NEJM’s authors are absolutely right – but they’re dead-wrong in concluding there’s no place for lean in “many vital aspects of medicine.” As with our R&D research, we’ve found that lean deployment in hospital settings minimizes unnecessary variation that comes from care providers, not patients. In fact, it frees up time and effort to cater to the necessary variability in a population diverse in its illnesses, economic backgrounds, languages and more.

 

Huddle2

As an example, in a recent study, my fellow researchers that include physician leaders and I spent years looking at Ohio State University’s Wexner Medical Center, specifically a lean deployment effort in its kidney transplant discharge process (see right picture). Transplant recipients after discharge must drink at least three liters of fluid a day – failure to do so can spike creatine levels, elevating blood pressure and increasing the likelihood of readmission. This is an evidenced based fact well-grounded in the medical literature7. During our roundings, we found variations in how these instructions were delivered to the patients: One nurse, for example, nonspecifically recommended drinking “a lot of fluids” while another suggested 100 ounces. Interestingly, we found nurses varying instructions between patients and patients who received different instructions from more than one nurse. These nurses weren’t being intentionally deceptive but their inconsistencies confused patients as they took in a tremendous overload of instructions.

Overhauled through the lean deployment via standard work design, nurses in our medical center now clearly explain the specific volume of fluid, use a jug to visually illustrate, and discuss the consequences of not following the instructions. Preliminary findings show this approach soothes patients’ anxiety levels and has reduced the chances of readmissions in the first month after transplant.

This isn’t just a lean approach to a problem – it’s a smart lean approach. And in an environment that, yes, isn’t cars on an assembly line that matters more than ever.

References:

  • Chandrasekaran, A., Linderman, K., Sting, F., Benner, M. 2016, Managing R&D Project Shifts in High-Tech Organizations: Multi-Method Study. Forthcoming at Production and Operations Management
  •  Chandrasekaran, A., Linderman, K., and Schroeder, R.G. 2012. Antecedents to Ambidexterity Competency in High Technology Organizations. Journal of Operations Management, 30(1-2) 134-151.
  •  Chandrasekaran, A., Senot, C., Boyer, K. 2012. Process Management Impact on Clinical and Experiential Quality: Managing Tensions between Safe and Patient-Centered Healthcare. Manufacturing and Service Operations Management, 14(4) 548-566
  •  Senot, C., Chandrasekaran. A. 2015. What has the highest impact on Readmissions Rates? Harvard Business Review (A joint collaboration between HBR-NEJM)
  •  Senot, C., Chandrasekaran, A., Ward, P., Tucker, A., Moffat-Bruce, S (2016) The Impact of Combining Conformance Quality and Experiential Quality on Readmissions and Cost Performance. Forthcoming Management Science
  •  Chandrasekaran, A., Anand. G., Sharma L., Ward, P., Henry M., Pesavento T., Moffatt-Bruce. The relationship between Patient Anxiety, Quality of Discharge Instructions and 30-day Readmission Rates for Kidney Transplant Recipients. (Findings presented at the American Surgical Congress, Feb 2016)
  • Gordon, E.J., Prohaska, T., Gallant, M., Siminoff, L., (2009). Self-care strategies and barriers among kidney transplant recipients: a qualitative study. Chronic Illness. 5(2) 75-91.